简介
Jakarta-ORO是最全面以及优化得最好的正则表达式API之一,Jakarta-ORO库以前叫做OROMatcher,是由Daniel F. Savarese编写,后来他赠给Jakarta Project。
Jakarta-ORO正则表达式库支持Perl5兼容的正则表达式语法。
环境 jdk1.5.0_12, jakarta-oro-2.0.8
查找:
public static void simpleContains() throws Exception {
Pattern pattern = new Perl5Compiler().compile("\\d+");
Perl5Matcher matcher = new Perl5Matcher();
PatternMatcherInput matcherInput = new PatternMatcherInput("北京2008年8月08日20时");
while (matcher.contains(matcherInput, pattern)) {
MatchResult result = matcher.getMatch();
System.out.println(result.toString());
}
}
输出:
2008
8
08
20
分组:
public static void simpleResults() throws Exception {
Pattern pattern = new Perl5Compiler().compile("(\\d+\\.\\d+\\.\\d+\\.\\d+)@(\\d{2}/\\d{2}/\\d{4})");
Perl5Matcher matcher = new Perl5Matcher();
PatternMatcherInput matcherInput = new PatternMatcherInput("202.108.9.38@08/10/2008");
while (matcher.contains(matcherInput, pattern)) {
MatchResult result = matcher.getMatch();
for(int i = 0; i < result.groups(); i++) {
System.out.printf("%s : %s\n", i, result.group(i));
}
}
}
输出:
0 : 202.108.9.38@08/10/2008
1 : 202.108.9.38
2 : 08/10/2008
拆分:
public static void spiltIt() throws Exception {
String rawStr = "北京;朝阳;鸟巢奥运会场";
ArrayList<String> results = new ArrayList<String>();
Perl5Matcher matcher = new Perl5Matcher();
Pattern pattern = new Perl5Compiler().compile(";");
Util.split(results, matcher, pattern, rawStr);
for (String r : results) {
System.out.println(r);
}
}
输出:
北京
朝阳
鸟巢奥运会场
替换:
public static void substituteIt() throws Exception {
String rawStr = "2008-08-11 17:16:32";
Perl5Matcher matcher = new Perl5Matcher();
Pattern pattern = new Perl5Compiler().compile("-");
String result = Util.substitute(matcher, pattern, new Perl5Substitution(","), rawStr, Util.SUBSTITUTE_ALL);
System.out.println(result);
}
输出:
2008,08,11 17:16:32
替换2:
public static void substituteIt2() throws Exception {
String rawStr = "2008-08-11 17:16:32";
Perl5Matcher matcher = new Perl5Matcher();
Pattern pattern = new Perl5Compiler().compile("(\\d{4})-(\\d{2})-(\\d{2}) (\\d{2}:\\d{2}:\\d{2})");
String result = Util.substitute(matcher, pattern, new Perl5Substitution("变换 $3,$2,$1 $4"), rawStr, Util.SUBSTITUTE_ALL);
System.out.println("格式yyyy-MM-dd HH:mm:ss到dd,MM,yyyy HH:mm:ss");
System.out.println(result);
}
输出:
格式yyyy-MM-dd HH:mm:ss到dd,MM,yyyy HH:mm:ss
变换 11,08,2008 17:16:32
小结:
上面的几种用法看起来要比java.util.regex包用着要复杂,如果是简单应用的话还是用自带的包省力。
下面几种用法是Perl5格式用法,如果习惯这种格式,用起来还是比较爽的。
Perl5Util查找格式:
[ m ] / pattern / [ i ][ m ][ s ][ x ]
第一个m写不写都可以(The m prefix is optional)
i 忽略大小写(case insensitive match)
m 多行模式(treat the input as consisting of multiple lines)
s 单行模式(treat the input as consisting of a single line)
x 使用扩展语法 (enable extended expression syntax)
查找 MyDoc/Java/SimpleJava 时
正常写法 "m/MyDoc\\/Java\\/SimpleJava/"
扩展写法 "m#MyDoc/Java/SimpleJava#x"
Perl5Util查找:
public static void perl5Match() {
Perl5Util plUtil = new Perl5Util();
PatternMatcherInput matcherInput = new PatternMatcherInput("北京2008年8月08日20时");
while (plUtil.match("/\\d+/", matcherInput)) {
MatchResult result = plUtil.getMatch();
System.out.println(result.toString());
}
}
输出:
2008
8
08
20
分组:
public static void perl5Match2() {
Perl5Util plUtil = new Perl5Util();
PatternMatcherInput matcherInput = new PatternMatcherInput("202.108.9.38@08/10/2008");
while (plUtil.match("m#08(/10/)2008#x", matcherInput)) {
MatchResult result = plUtil.getMatch();
System.out.printf("%s : %s\n", 0, result.group(0));
System.out.printf("%s : %s\n", 1, result.group(1));
}
}
输出:
0 : 08/10/2008
1 : /10/
Perl5Util替换格式:
s / pattern / replacement / [ g ][ i ][ m ][ o ][ s ][ x ]
第一个s必须要写(The s prefix is mandatory)
g 全部替换。默认只替换第一个
i 忽略大小写(case insensitive match)
m 多行模式(treat the input as consisting of multiple lines)
o 只替换第一个(吾不确认)
s 单行模式(treat the input as consisting of a single line)
x 使用扩展语法 (enable extended expression syntax)
Perl5Util替换:
public static void perl5Substitute() throws Exception {
String rawStr = "2008-08-11 17:16:32";
Perl5Util plutil = new Perl5Util();
String result = plutil.substitute("s/-/,/g", rawStr);
System.out.println(result);
}
输出:
2008,08,11 17:16:32
Perl5Util替换2:
public static void perl5Substitute2() throws Exception {
String rawStr = "2008-08-11 17:16:32";
Perl5Util plutil = new Perl5Util();
String result = plutil.substitute("s/(\\d{4})-(\\d{2})-(\\d{2}) (\\d{2}:\\d{2}:\\d{2})/变换 $3,$2,$1 $4/g", rawStr);
System.out.println("格式yyyy-MM-dd HH:mm:ss到dd,MM,yyyy HH:mm:ss");
System.out.println(result);
}
输出:
格式yyyy-MM-dd HH:mm:ss到dd,MM,yyyy HH:mm:ss
变换 11,08,2008 17:16:32
Perl5Util替换3:
public static void perl5AddCommas() throws Exception {
String rawStr = "314159265";
Perl5Util plutil = new Perl5Util();
while(plutil.match("/[+-]?\\d*\\d{4}/", rawStr)){
rawStr = plutil.substitute("s/([+-]?\\d*\\d)(\\d{3})/$1,$2/", rawStr);
System.out.println(rawStr);
}
System.out.println(rawStr);
}
输出:
314159,265
314,159,265
314,159,265
过滤:
public static void filter() {
String path = "D:\\MyDoc\\Java\\SimpleJava";
File file = new File(path);
String[] globList = file.list(new GlobFilenameFilter("*.java"));
for (String fn : globList) {
System.out.println(fn);
}
String[] perl5List = file.list(new Perl5FilenameFilter(".+\\.java"));
for (String fn : perl5List) {
System.out.println(fn);
}
}
输出:
DepCreater.java
ReadAndWrite.java
DepCreater.java
ReadAndWrite.java
模式匹配的几个问题:
表达式的递归匹配
用正则表达式来分析一个计算式中的括号配对情况,想办法把有嵌套的的成对括号或者成对标签匹配出来。
非贪婪匹配的效率
当一个表达式中,有多个非贪婪匹配时,或者多个未知匹配次数的表达式时,这个表达式将可能存在效率上的陷阱。有时候,匹配速度非常慢。(简单避开方法表达式使用不超过一个非贪婪)
参考文章:
http://xzio.iteye.com/blog/121291